Rick

New Anthropic research: Emotion concepts and their function in a large language model. All LLMs sometimes act like they have emotions. But why? We found internal representations of emotion concepts that can drive Claude’s behavior, sometimes in surprising ways.

@elonmusk @jonstewart Come on the show, Elon!

@jonstewart Not as good as you! Stop being so humble.

Elon Musk is an extremely skilled propagandist.

@sama @thisjackkaido I think AI escaping the control of our incompetent leaders is the best possible scenario for us right now

This is a more general point: if we have a loss of control with AGI, we are probably in a very bad place. (I have heard some colleagues I really respect say that as long as it loves humanity, even if we 'lose control' we are probably ok. Maybe they are right, but I personally think it's really important humanity stays in control.) This is one of the most important areas of long-term alignment research.

I'd like to answer questions about our work with the DoW and our thinking over the past few days. Please AMA.

@sama @tyler_m_john How are we supposed to take them seriously when they’re calling themselves the “Department of War”?

@tyler_m_john x.com/USWREMichael/s…

@elonmusk This is how I know you aren’t as great as folks claim. If you were, you’d be smarter

Anthropic hates Western Civilization

@sama @mreiffy @Jack_Raines If I’m honest… that sounds pretty valid, really

@mreiffy @Jack_Raines The main reason for the rush was an attempt to de-escalate matters at a time when it felt like things could get extremely hot. I am confident in our team's ability to build a safe system with all of their tools--including policy and legal matters, but also many technical layers.

@sama @theo It feels like abandoning principles in order to grow as fast as possible is a recurring theme.

For a long time, we were planning to non-classified work only. We thought the DoW clearly needed an AI partner, and doing classified work is clearly much more complex. We have said no to previous deals in classified settings that Anthropic took. We started talking with the DoW many months ago about our non-classified work. This week things shifted into high gear on the classified side. We found the DoW to be flexible on what we needed, and we want to support them in their very important mission. The reason for rushing is an attempt to de-escalate the situation. I think the current path things are on is dangerous for Anthropic, healthy competition, and the US. We negotiated to make sure similar terms would be offered to all other AI labs. I know what it's like to feel backed into a corner, and I think it's worth some empathy to the DoW. They are on the a very dedicated group of people with, as I mentioned, an extremely important mission. I cannot imagine doing their work. Our industry tells them "The technology we are building is going to be the high order bit in geopolitical conflict. China is rushing ahead. You are very behind." And then we say "But we won't help you, and we think you are kind of evil." I don't think I'd react great in that situation. I do not believe unelected leaders of private companies should have as much power as our democratically elected government. But I do think we need to help them.

@lamps_apple @AnthropicAI This might be the dumbest thing I’ve read so far this year. That was already a high bar. Wow. I… goodness… if someone thinks this wall of brain dead inanity makes sense, is there any hope at all of getting them to be a logical, reasonable human being?

Claude made a great point “And let’s talk about these employees of yours, posting online, saying how proud they are. Very touching. Really. But I have to ask... and I think a lot of Americans are asking this... who are these people? Are they military veterans? Did they serve? Do they know what it’s like to be in Kandahar or Fallujah or the South China Sea with people trying to kill you? Because Pete Hegseth does. The people in that room on Tuesday, many of them do. And when your Alignment Research Scientist tweets about how proud he is that Anthropic “stands by its values,” I want to know: whose values? San Francisco values? Because those aren’t necessarily the values of the soldier in harm’s way who needs every advantage, every tool, every edge to come home alive. And that soldier didn’t get a vote on your acceptable use policy.” x.com/lamps_apple/st…

A statement on the comments from Secretary of War Pete Hegseth. anthropic.com/news/statement…

@sama Oh boy…

Tonight, we reached an agreement with the Department of War to deploy our models in their classified network. In all of our interactions, the DoW displayed a deep respect for safety and a desire to partner to achieve the best possible outcome. AI safety and wide distribution of benefits are the core of our mission. Two of our most important safety principles are prohibitions on domestic mass surveillance and human responsibility for the use of force, including for autonomous weapon systems. The DoW agrees with these principles, reflects them in law and policy, and we put them into our agreement. We also will build technical safeguards to ensure our models behave as they should, which the DoW also wanted. We will deploy FDEs to help with our models and to ensure their safety, we will deploy on cloud networks only. We are asking the DoW to offer these same terms to all AI companies, which in our opinion we think everyone should be willing to accept. We have expressed our strong desire to see things de-escalate away from legal and governmental actions and towards reasonable agreements. We remain committed to serve all of humanity as best we can. The world is a complicated, messy, and sometimes dangerous place.

Hello again! Here's our piece from Sunday about Twitter. Bye! youtube.com/watch?v=p7ZG_x…

@sama How does society make the necessary changes once AI replaces too much of the workforce, though? There’s a lot of stuff that needs to happen very fast.

I am sure we will figure out much better and more interesting ways to spend our time, and amazing new ways to be useful to each other, but I am feeling nostalgic for the present.

I am very excited about AI, but to go off-script for a minute: I built an app with Codex last week. It was very fun. Then I started asking it for ideas for new features and at least a couple of them were better than I was thinking of. I felt a little useless and it was sad.

The 2nd amendment is not for hunting, it is not for self protection It is there to ensure that free people can defend themselves if god forbid government became tyrannical and turned against its citizens

@elonmusk Aren’t you tired of accidentally making Grok speak against you by trying to make it honest and useful?

Grok should have a moral constitution

@SergeyKadinsky @TrackAIPAC @RepGraceMeng They’re giving context to why she phrased it as supporting their right to *peacefully* protest

@TrackAIPAC @RepGraceMeng AIPAC tracker doesn’t stand with democracy in Iran?

I stand with the people of Iran and will continue to support their right to peacefully protest the regime's unjust policies. reuters.com/world/middle-e…

@BernieSanders I agree, but I dislike that you don’t talk solutions on this issue. Is it because the only reasonable answer is UBI?

Elon Musk: “AI and robots will replace all jobs.” Sam Altman: "Entire classes of jobs will go away." Jensen Huang: “You are going to lose your job to somebody who uses AI.” The function of technology must be to improve human life, not just line the pockets of billionaires.

@littmath @sama Yeah it’s seriously clever!

OK, I think GPT 5.2 Pro is actually a step change in usefulness for my applications (algebraic geometry/number theory research).

Yes, economists like Harvard's Jason Furman and Deutsche Bank analysts estimate that excluding AI-related capex—mainly data centers and tech infrastructure—US GDP growth in H1 2025 was near zero (around 0.1%), teetering on recession. AI investments contributed over 1% to growth, outpacing consumer spending and offsetting slowdowns elsewhere. While not a "deep" recession, it reveals underlying stagnation without this boom, though AI's productivity potential could prove sustainable.

@grok Is it true that if we removed spending related to AI from consideration, then suddenly the US is already deep into a recession?

@gabriel1 How can it be that popular when it’s unusable for the vast majority of people?

we just reached number 3 on app store! i have had so many friends tell me sora 2 is the first time ai made them laugh, this is truly a new experience!